Google AI Studio

Google AI Studio is a web-based development environment used for rapid prototyping and experimentation with Google's generative AI models, primarily the Gemini family. Its main purposes are to allow users to quickly test and refine prompts (including multimodal inputs like images), adjust model settings, and instantly export production-ready code via the Gemini API.

Key Features

Please note that access to some features may be limited based on the version of AI Studio you're using (see Eligibility & Access).

AI Studio enables you to quickly test out different prompts, inputs (including text, images, and other media), and model settings to see how the AI responds.

AI Studio provides direct access to Google's powerful models, such as Gemini Flash and Gemini Pro, as well as models for specific tasks like image and video generation (e.g., Imagen and Veo).

Easily refine your prompts, set system instructions (to define the model's persona or rules), and adjust parameters like "temperature" (to control creativity/randomness) without needing to write code.

The platform supports uploading various media types (images, documents, audio, and sometimes video) to test the models' ability to understand and respond to different kinds of data.

Once you have a prompt and model configuration that works, you can click "Get Code" to instantly generate a working code snippet (often in Python, Node.js, etc.) that you can use to integrate that functionality into your own application via the Gemini API

Exported code may be required to undergo evaluation for compliance with NYU IT standards, including cybersecurity, privacy, and digial accessibility.

Eligibility & Access

There are two tiers of AI Studio available. The differences between the free version of Google AI Studio and the paid tier primarily relates to the Gemini API quotas, data privacy, and access to advanced features needed for production-scale applications.

  • Access is limited to those who have an approved Google Cloud Project and is activated as part of the overall project approval process. 
  • AI Studio comes with a set number of credits. Projects that exceed this quota may incur an additional fee. This is be estimated beforehand as part of the Google Cloud Project approval process.

To use the paid service:

  1. Log in at https://aistudio.google.com with your NYU NetID username and password.
  2. Once approved for Paid Tier access, click Get API key.
  3. On this new page, click Create API key on the upper right. You will be prompted to create a project or import one.
  4. Once a project is created or imported, you can then select it from the drop down list and the API key will be registered.

Key Differences Between Tiers

  • Scaling and Performance: The Free Tier is an appetizer for testing—you will quickly hit rate limits if you try to build a real application used by multiple people. The Paid Tier is essential for production workloads due to the massively increased quotas.
  • Privacy and Training: The Paid Tier offers a crucial privacy benefit: your data is not used for model training, which is a necessity for most commercial or sensitive applications.
  • Advanced Tools: Features designed to optimize cost and performance at scale (like Caching and Batching) are only available in the Paid Tier.

Enterprise Use of Applications

All applications developed with AI Studio will be subject to evaluation by NYU IT, during which an implementation strategy for proposed applications will be explored.

Contact NYU IT